Do Improvements in Motivational Language Predict Alcohol Use in Motivational Interviewing? Ambivalence Matters

Objective: Motivational Interviewing (MI) is described as a method for improving clinical outcomes by reducing client ambivalence. If this is true, MI's focus on improving clients' motivational language should be most useful for clients with ambivalence about change and less valuable for those who are ready to implement new behaviors or are opposed to change. To address this hypothesis and potentially add precision to MI delivery in clinical settings, we tested whether the relationship between clients' in-session motivational language and posttreatment alcohol use depended on their baseline motivation to change. Method: Client speech from 149 sessions from Project MATCH were analyzed. A cluster analysis of the percent change talk during the first decile of the session identified three motivational groups: opposed, ambivalent, and ready. The change in percent change talk (C-PCT) across the session was calculated for each group. Zero-inflated negative binomial analysis was used to test whether the effect of C-PCT on end-of-treatment drinking varied between motivational groups. Results: The count part of the model revealed a significant interaction between C-PCT and membership in the ambivalent group (b = -17.710, 95% CI [-25.775, -9.645], p < .001), only for those who received MI. Favorable C-PCT was associated with less drinking (b = -15.735, p = .004). Only baseline drinking was a significant predictor of abstinence at follow-up (b = .032, 95% CI [0.012, 0.051], p = .001). Conclusion: A putative MI mechanism-improved client motivational language-appears most important for clients who express ambivalence in the opening minutes of the session, with minimal value for those who do not.